Table 2 Ovarian & Breast Cancer Risk Study Characteristics and Results

Breast Cancer Risk

Patient Characteristics

Study Design, Data Source

Lifestyle Factor

Measurement Method

Results

3067 women with BRCA pathogenic germline gene variants

2498 BRCA1

569 BRCA2

Prospective cohort, data from 78 participating centers in 12 countries

Average 5.4-year follow-up

Dietary habits- alcohol

Standardized questionnaire including questions related to family and personal history of cancer, medical and reproductive history, and the following questions related to alcohol consumption: Current consumption, age at first and last use, average number of drinks per week, type of alcohol consumed.

Baseline questionnaire completed at time of clinic appointment and follow-up questionnaires completed every 2 years thereafter

259 incident cases observed.

Significant relationships were not observed between breast cancer risk and the following alcohol variables in adjusted models*: ever use of alcohol, cumulative consumption, age at first use, alcohol use by the first full-term birth.

Significant relationships were not observed between ever or current use of alcohol and breast cancer risk by menopausal status, pathogenic gene variant, and age of breast cancer diagnosis among cases.

*Data adjusted for age at baseline, BRCA1/2 pathogenic germline gene variant, age at menarche, oral contraceptive use, history of breast feeding, mean parity, oophorectomy status, and country of residence.

1925 matched case-control pairs of women with BRCA1/2 pathogenic germline gene variants

1480 BRCA1

445 BRCA2

Case-Control, data from 54 centers in 8 countries

Dietary habits- alcohol

Standardized questionnaire with questions related to alcohol consumption: if consume alcohol, number of drinks per week.

Drinks consumed per week and breast cancer risk in women withBRCA1pathogenic germline gene variants*:

BRCA1

--none**: 1.00

--0-3: OR, 0.77 (0.67,0.94)

--4-9: OR, 0.98 (0.73,1.32)

-- ≥ 10: OR, 0.55 (0.33,0.91)

p-trend = 0.03

Type of alcohol consumed per week and breast cancer risk among women withBRCA1pathogenic germline gene variants*:

exclusive wine consumers

--none**: 1.00

--0-3: OR, 0.62 (0.45,0.87)

--4-9: OR, 0.82 (0.41,1.67)

-- ≥ 10: OR, 0.39 (0.11,1.45)

p-trend = 0.01

other alcohol types (beer and spirits)

--none**: 1.00

--0-3: OR, 0.62 (0.43,0.91)

--4-9: OR, 1.07 (0.40,2.85)

-- ≥ 10: OR, 0.70 (0.13,3.75)

p-trend = 0.01

Significant associations not observed in women with BRCA2 pathogenic germline gene variants for any alcohol variables.

*Data adjusted for ethnicity, menopause, oral contraceptive use, hormone-replacement therapy use, smoking status, history of oophorectomy, BMI, and parity.

**Individuals who reported that they did not currently consume alcoholic beverages

857 breast cancer cases diagnosed within the last 10 years of data collection

10 cases with BRCA1 pathogenic germline gene variant

33 cases with BRCA2 pathogenic germline gene variant

814 cases without BRCA pathogenic germline gene variant

Case-only, data from Centre Hospitalier de L’Universite de Montreal

Dietary habits- alcohol

Interviewer administered food frequency questionnaire developed by the NCI of Canada. Questionnaire inquired about alcohol consumption in the year prior to breast cancer diagnosis

Also completed questionnaire related to other lifestyle factors: ethnicity, family history, reproductive and medical history, menopausal status, smoking habits, oral contraceptive use, hormone replacement therapy use (data not shown in this table)

Average time between diagnosis and interview was 3.1 years.

10/10 (100%) women with BRCA1 pathogenic germline gene variants consumed alcohol within the year prior to breast cancer diagnosis.

30/33 (90.9%) women with BRCA2 pathogenic germline gene variants consumed alcohol within the year prior to breast cancer diagnosis.

Alcohol consumption among women with breast cancer andBRCA1/2pathogenic germline gene variants compared to women with breast cancer withoutBRCA1/2pathogenic germline gene variant*^,**:

BRCA1

--total alcohol (3 drinks/week): COR, 0.79 (0.22,2.83)

--wine (2 drinks/week): COR, 0.38 (0.08,1.81)

--other alcohol (0.33 drinks/week): COR, 2.49 (0.64,9.73)

BRCA2

--total alcohol (3 drinks/week): COR, 1.99 (0.96,4.11)

--wine (2 drinks/week): COR, 1.58 (0.78,3.17)

--other alcohol (0.33 drinks/week): COR, 2.15 (1.03,4.50)

*Data adjusted for age at diagnosis

**Drinks per week dichotomized by case median

1337 women with BRCA pathogenic germline gene variants

499 women with breast cancer and BRCA pathogenic germline gene variant

− 332 BRCA1

−167 BRCA2

838 women without breast cancer but with BRCA pathogenic germline gene variant

− 531 BRCA1

−307 BRCA2

Case-Control, data from French National BRCA 1/2 Carrier Cohort (GENEPSO)

Dietary habits- alcohol

Standardized questionnaire administered by mail inquiring about reproductive factors, tobacco use, alcohol consumption at age 20, and history of chest x-ray exposure

Among women withBRCA1pathogenic germline gene variant*:

- When alcohol use was stratified by tobacco use (ever vs never smoker) there were no significant interactions observed (p > 0.05).

-When tobacco use was stratified by alcohol use (ever vs never use of alcohol) the only significant interactions observed were among women who reported never drinking alcohol.

Among women withBRCA2pathogenic germline gene variant*:

Ever use

--No: 1.00

--Yes: HR, 1.21 (0.68,2.15)

Consumed > 5 glasses per week at age 20

--No: 1.00

--Yes: HR, 1.78 (0.97,3.27)

-There were no significant interactions between alcohol and tobacco use (p = 0.75). Therefore, analysis for tobacco and alcohol use were not stratified among women with BRCA2 pathogenic germline gene variants as it was for women with BRCA1 pathogenic germline gene variant.

*Data adjusted for parity, menopausal status, gene, smoking history, number of years of smoking interruption

804 women with BRCA pathogenic germline gene variants

323 women with breast cancer

−195 BRCA1

− 128 BRCA2

481 women without breast cancer

− 302 BRCA 1

− 179 BRCA 2

Case-Control, data from six research institutions in USA, Canada, and Australia who were part of Breast Cancer Family Registry, and both the Kathleen Cuningham Foundation Consortium for Research into Familial Breast Cancer in Australia, and the Ontario Cancer Genetics Network in Canada

Dietary habits- alcohol

Risk factor questionnaire including questions related to alcohol consumption

Alcohol consumption and breast cancer risk in women with theBRCA1pathogenic germline gene variant*,**:

Ever use

--No: 1.00

--Yes: OR, 1.06 (0.73,1.52)

Current use

--No: 1.00

--Yes: OR, 0.96 (0.67,1.37)

Years of drinking

--Nonusers: 1.00

--1-29: OR, 1.07 (0.64,1.76)

-- > 29: OR, 0.93 (0.62,1.39)

--Trend per 10 years of drinking: OR, 0.98 (p = 0.5)

Daily alcohol intake (g/d)

--Nonusers: 1.00

--1-4: OR, 0.63 (0.34,1.18)

-- > 4: OR, 1.14 (0.77,1.69)

--Trend per 10 g: OR, 1.02 (p = 0.4)

Alcohol consumption and breast cancer risk in women with theBRCA2pathogenic germline gene variant*,**:

Ever use

--No: 1.00

--Yes: OR, 0.66 (0.45,0.97)

Current use

--No: 1.00

--Yes: OR, 1.11 (0.76,1.63)

Years of drinking

--Nonusers: 1.00

--1-29: OR, 0.40 (0.21,1.74)

-- > 29: OR, 0.89 (0.59,1.34)

--Trend per 10 years of drinking: OR, 1.02 (p = 0.4)

Daily alcohol intake (g/d)

--Nonusers: 1.00

--1-4: OR, 0.41 (0.22,0.77)

-- > 4: OR, 0.79 (0.52,1.18)

--Trend per 10 g: OR, 1.00 (p = 0.9)

Significant differences in breast cancer risk was not observed by alcohol type (i.e. wine, beer, liquor).

*Data adjusted for age (as a continuous variable), family history (number of first degree relatives with history of breast or ovarian cancer), smoking status, and number of full-term pregnancies.

**Stratified by age (<  40 years and > 40–49 years) and study sites.

1381 female breast cancer cases

283 women with breast cancer and BRCA 1 pathogenic germline gene variant

204 women with breast cancer and BRCA 2 pathogenic germline gene variant

894 sporadic breast cancer cases

Case-Only, data from the Genetic and Environmental Modifiers of BRCA 1 and BRCA 2 pathogenic germline gene variants Study (GEMS) Cases were identified either prospectively or retrospectively pending on which center was collecting data

Dietary habits- alcohol

Weight status

Risk factor questionnaire that inquired about demographic information, medical and reproductive history, use of oral contraceptives, smoking status, alcohol use, and weight history

Prospective enrollment of breast cancer cases used either the GEMS questionnaire or included a supplement to a pre-existing questionnaire to capture lifestyle data not included in the original questionnaire but included in the GEMS questionnaire

Retrospective enrollment of breast cancer cases used a similar risk factor questionnaire but not the specific GEMS questionnaire

Alcohol use and breast cancer risk in women with breast cancer andBRCA1pathogenic germline gene variant compared to women with breast cancer withoutBRCA1pathogenic germline gene variant*:

Never use: 1.00

Ever use: IRR, 0.65 (0.48,0.90)

Weight history and breast cancer risk in women with breast cancer andBRCA1pathogenic germline gene variant compared to women with breast cancer withoutBRCA1pathogenic germline gene variant*:

BMI (kg/m²) one year before diagnosis

-- < 25: IRR, 1.00

-- ≥ 25 to < 30: IRR, 0.99 (0.65,1.51)

-- ≥ 30: IRR, 1.38 (0.89,2.13)

BMI (kg/m²) at age 18

-- < 25: IRR, 1.00

-- ≥ 25 to < 30: IRR, 0.76 (0.48,1.20)

-- ≥ 30: IRR, 1.15 (0.68,1.94)

Alcohol use and breast cancer risk in women with breast cancer andBRCA2pathogenic germline gene variant compared to women with breast cancer withoutBRCA2pathogenic germline gene variant*:

Never use: 1.00

Ever use: IRR, 0.80 (0.55,1.16)

Weight history and breast cancer risk in women with breast cancer andBRCA2pathogenic germline gene variant compared to women with breast cancer withoutBRCA2pathogenic germline gene variant*:

BMI (kg/m²) one year before diagnosis

-- < 25: IRR, 1.00

-- ≥ 25 to < 30: IRR, 1.09 (0.70,1.70)

-- ≥ 30: IRR, 1.15 (0.67,1.90)

BMI (kg/m²) at age 18

-- < 25: IRR, 1.00

-- ≥ 25 to < 30: IRR, 0.81 (0.50,1.31)

-- ≥ 30: IRR, 0.68 (0.33,1.38)

*Data adjusted for age and center.

348 matched case-control pairs with BRCA1 pathogenic germline gene variant

Case-control, data from International Hereditary Cancer Center in Szczecin or elsewhere in Poland

Dietary habits- coffee

Standardized questionnaire that inquired about reproductive and medical history, smoking history, oral contraceptive use, and coffee consumption.

845 matched case-control pairs

652 pairs with BRCA1

193 pairs with BRCA2

Cases were diagnosed with breast cancer as their first or only cancer

Case-Control, data from 40 centers in 4 countries

Dietary habits- coffee

Standardized questionnaire that inquired about demographic information, ethnicity, parity, family history, reproductive and medical history, use of oral contraceptives, smoking history, alcohol consumption and coffee consumption.

-questions related to caffeinated and decaffeinated coffee consumption include: ever use, current use, age when started drinking coffee, age when stopped drinking coffee, average daily coffee consumption

7.8 years on average elapsed from diagnosis date to questionnaire administration

Caffeinated coffee consumption and breast cancer risk*:

--0 cups/day: 1.00

--1-3 cups/day: OR, 0.90 (0.72,1.12)

--4-5 cups/day: OR, 0.75 (0.47,1.19)

-- ≥ 6 cups/day: OR, 0.31 (0.13,0.71)

p-trend = 0.02

Decaffeinated coffee consumption and breast cancer risk*:

--0 cups/day: 1.00

--1-3 cups/day: OR, 0.99 (0.72,1.36)

--4-5 cups/day: OR, 1.14 (0.30,4.31)

-- ≥ 6 cups/day: NA

p-trend = 1.00

Total coffee consumption (caffeinated + decaffeinated) and breast cancer risk*:

--0 cups/day: 1.00

--1-3 cups/day: OR, 0.89 (0.70,1.13)

--4-5 cups/day: OR, 0.73 (0.48,1.10)

-- ≥ 6 cups/day: OR, 0.51 (0.26,0.98)

p-trend = 0.03

Caffeinated coffee consumption and breast cancer risk byBRCA1/2pathogenic germline gene variants*:

BRCA1

--0 cups/day: 1.00

--1-3 cups/day: OR, 0.82 (0.64,1.06)

--4-5 cups/day: OR, 0.67 (0.39,1.16)

-- ≥ 6 cups/day: OR, 0.25 (0.09,0.71)

p-trend = 0.009

BRCA2

--0 cups/day: 1.00

--1-3 cups/day: OR, 1.26 (0.78,2.08)

--4-5 cups/day: OR, 1.17 (0.48,2.83)

-- ≥ 6 cups/day: OR, 0.40 (0.09,1.73)

p-trend = 0.84

*Data adjusted for parity, smoking status, oral contraceptive use, alcohol consumption and BMI at age 30.

491 women with BRCA1/2 pathogenic germline gene variant

370 cases with breast cancer

1789 women without pathogenic germline gene variants

− 1632 cases with breast cancer

Included in an analysis of all breast cancer cases regardless of pathogenic germline gene variant status

(this data is not reported in this table)

Retrospective cohort, data from KOHBRA (Korean Hereditary Breast Cancer Study)

Dietary habits- food intake

Validated food frequency questionnaire developed by the Korean National Institutes of Health

Dietary intake divided into quartiles. Intake of the following food items was assessed: vegetables, fruit, meat, seafood, soybean products.

Significant associations were not observed in women with BRCA1/2 pathogenic gene variants for vegetable, fruit and seafood intake in and breast cancer risk.

Meat and soybean product intake and breast cancer risk among women withBRCA1/2pathogenic germline gene variants carriers combined*:

Meat (number of food items)

--Q1 (0): 1.00

--Q2 (1): HR, 1.03 (0.64,1.68)

--Q3 (2): HR, 1.29 (0.77,2.17)

--Q4 (3–10): HR, 1.97 (1.13, 3.44)

p-trend = 0.026

Soybean products (number of food items)

--Q1 (0–1): 1.00

--Q2 (2): HR, 1.09 (0.68,1.76)

--Q3 (3): HR, 0.72 (0.45,1.14)

--Q4 (4–5): HR, 0.39 (0.19, 0.79)

p-trend = 0.005

Meat and soybean product intake and breast cancer risk among women withBRCA2pathogenic germline gene variant*:

Meat (number of food items)

--Q1 (0): 1.00

--Q2 (1): HR, 0.83 (0.42,1.64)

--Q3 (2): HR, 1.16 (0.57,2.37)

--Q4 (3–10): HR, 2.48 (1.26, 4.89)

p-trend = 0.027

Soybean products (number of food items)

--Q1 (0–1): 1.00

--Q2 (2): HR, 1.41 (0.75,2.65)

--Q3 (3): HR, 0.76 (0.40,1.44)

--Q4 (4–5): HR, 0.38 (0.16, 0.93)

p-trend = 0.005

Significant associations were not observed in women with BRCA1 pathogenic germline gene variant for meat and soybean intake and breast cancer risk

*Data adjusted for menarche, caloric intake, years of education, smoking history, alcohol intake history, exercise habits, and parity.

89 cases with BRCA1 or BRCA2 pathogenic germline gene variants

48 controls with BRCA1 or BRCA2 pathogenic germline gene variants

46 controls who did not have BRCA1/2 pathogenic germline gene variants

Case-Control, data from 80 French Canadian families

Dietary habits- diet quality

Validated food frequency questionnaire developed by the National Cancer Institute of Canada. The questionnaire covered the 1-year period prior to diagnosis for cases and the corresponding time period for controls. Included dietary habits, multivitamins, supplements and alcohol use.

Assessed dietary intake via the following diet-quality indexes:

-Alternative healthy eating index (AHEI)

-Diet quality index-revised (DQI-R)

-alternate Mediterranean diet index (aMED)

-Canadian healthy eating index (CHEI)

Also included general lifestyle questionnaire to collect baseline data on other lifestyle variables.

Logistic regression analysis was only conducted on diet quality variables.

The only significant differences between cases and controls were among the following variables (p < 0.05):

total energy intake (kcal/d)

--Cases: 2589 ± 1142

--Controls- BRCA 1/2 carriers: 2167 ± 830

--Controls- non-carriers: 2146 ± 720

age at maximum weight (years)

--Cases: 46.2 ± 13.3

--Controls- BRCA 1/2 carriers: 38.8 ± 13.3

--Controls- non-carriers: 41.5 ± 16.5

maximum BMI (kg/m²)

--Cases: 27.4 ± 5.6

--Controls- BRCA 1/2 carriers: 25.0 ± 4.5

--Controls- non-carriers: 26.8 ± 6.1

Diet quality and breast cancer risk when comparingBRCA1/2pathogenic germline gene variants cases and controls*:

DQI-R

--Q1: 1.00

--Q2: OR, 1.04 (0.43,2.52)

--Q3: OR, 0.35 (0.12,1.02)

p-trend = 0.034

CHEI

--Q1: 1.00

--Q2: OR, 0.42 (0.15,1.18)

--Q3: OR, 0.18 (0.05,0.68)

p-trend = 0.006

Diet quality and breast cancer risk when comparingBRCA1/2pathogenic germline gene variant cases and non-carrier controls*:

DQI-R

--Q1: 1.00

--Q2: OR, 0.54 (0.21,1.36)

--Q3: OR, 0.21 (0.07,0.62)

p-trend = 0.003

*Data adjusted for age, physical activity and total energy intake.

400 women with BRCA1/2 pathogenic germline gene variant

129 cases with breast cancer

Case-Control, data from 10 different centers in Canada

Dietary habits- nutrient intake

Folic acid B6

B12

Open-ended questionnaire collecting the following information about each supplement taken since age 18:

-type of supplement

-brand name of supplement

-weekly frequency of supplement use

-supplement dose

-duration of use

The following supplements were not significantly associated with breast cancer risk in adjusted models (with never use as the reference):

-Multivitamin, ever use

-Folic acid, ever use

-B6, 0.02 - ≤ 0.20 mg/d or > 0.20 mg/d

Ever use of prenatal supplement (with never use as reference):

OR, 0.57 (0.34, 0.95)*

OR, 0.60 (0.35, 1.02)**

Any folic-acid containing supplement (with never use as reference):

OR, 0.81 (0.50, 1.29)*

OR, 0.45 (0.25, 0.79)**

Total daily average of folic acid (with never use as reference):

8.56 - ≤ 89.29 mcg/d OR, 0.39 (0.19, 0.81)**

>  89.29 mcg/d OR, 0.54 (0.27, 1.10)**

Total daily average of B12 (with never use as reference):

0.02 - ≤ 0.34 mcg/d OR, 0.48 (0.24, 0.96)**

>  0.34 mcg/d OR, 0.61 (0.33, 1.12)**

Ever use of any folic-acid containing supplement, when assessed by BRCA mutation, revealed a significant association with breast cancer risk among women with BRCA1 pathogenic germline gene variants: OR, 0.30 (0.14,0.65)**

Ever use of any folic-acid containing supplement assessed by parity did not reveal significant association with breast cancer risk.

*Data adjusted for age and BRCA1/2 pathogenic germline gene variant

**Data adjusted for age, BRCA1/2 pathogenic germline gene variant, BMI, parity, alcohol consumption, smoking status

89 cases with BRCA1 or BRCA2 pathogenic germline gene variants

48 controls with BRCA1 or BRCA2 pathogenic germline gene variants

Case-Control, data from 80 French Canadian families

Dietary habits- nutrient intake

Weight change

Physical activity

Validated semi-quantitative food frequency questionnaire that covered the 1-year period prior to diagnosis for cases and the corresponding time period for controls

Lifestyle core questionnaire for physical activity, weight change, and other lifestyle factors such as smoking history, menopausal status, oral contraceptive use, medical and reproductive history

Physical activity information covered the 2-year period before diagnosis or interview for controls

Weight history information included height, current weight, weight at age 18 and 30.

Energy intake and breast cancer risk:

Total energy intake (kcal/d)*

--Q1 ≤ 1724: 1.00

--Q2 > 1724 and ≤ 2339: OR, 1.17 (0.44,3.13)

--Q3 > 2339: OR, 2.76 (1.10,7.02)

p-trend = 0.026

Significant associations were not observed for intake of the following in adjusted models: fat, protein, carbohydrates, poly-unsaturated fatty acids, mono-unsaturated fatty acids, saturated fatty acids, alcohol, beer, wine, spirits, vitamins C and E, fiber, folate, caffeine.

Weight change and breast cancer risk**:

Age at maximum BMI (years)

--Q1 ≤ 34: 1.00

--Q2 > 34 and ≤ 43: OR, 1.12 (0.41,3.05)

--Q3 > 43: OR, 2.90 (1.01,8.36)

p-trend = 0.043

Weight gain since age 18 (pounds)

--Q1 ≤ 12: 1.00

--Q2 > 12 and ≤ 35: OR, 3.63 (1.18,11.22)

--Q3 > 35: OR, 4.64 (1.52,14.12)

p-trend = 0.011

Weight gain since age 30 (pounds)

--Q1 ≤ 8: 1.00

--Q2 > 8 and ≤ 20: OR, 3.43(1.16,10.14)

--Q3 > 20: OR, 4.11 (1.46,11.56)

p-trend = 0.013

No significant association was observed between physical activity variables (i.e. weekly MET hours of moderate activity, weekly MET hours of vigorous activity, total weekly MET hours of physical activity) and breast cancer risk.

*Data adjusted for age, maximum lifetime BMI and physical activity

**Data adjusted for age, physical activity and total energy intake.

200 samples from women with BRCA1 rs80356932 & BRCA2 rs80359182 pathogenic germline gene variants

100 samples from women with breast cancer

Case-control, data from three hospitals in Pakistan: Jinnah Hospital, Fauji Foundation Hospital, and INMOL Hospital Lahore

Weight Status

BMI extracted from medical record

Breast cancer was most prevalent in women with obesity, per BMI (p = 0.002)

Normal weight (BMI 18.5–24.9 kg/m²) as reference

Underweight (BMI < 18.5 kg/m²): OR, 1.71 (0.41, 7.00)

Overweight (BMI 25.0–29.9 kg/m²): OR, 3.06 (1.36, 6.87)

Obese (BMI > 30 kg/m²): OR, 4.09 (1.91, 8.75)

1073 matched case-control pairs

797 pairs with BRCA1 pathogenic germline gene variants

276 pairs with BRCA2 pathogenic germline gene variants

Cases were diagnosed with breast cancer as their first or only cancer

Case-Control, data from 41 centers in 5 countries with research protocols including BRCA pathogenic germline gene variant status

Weight change

Standardized questionnaire that inquired about demographic information, ethnicity, parity, family history, reproductive and medical history, use of oral contraceptives, smoking history, weight at birth, age 18, 30 and 40, current weight and height

8.8 years on average elapsed from diagnosis date to questionnaire administration

Weight change between 18 and 30 years:

--loss of ≥10#: OR, 0.66 (0.46,0.93)

--loss of < 10# to gain of ≤10#: OR, 1.00

--gain of 10 to ≤20#: OR, 1.19 (0.96,1.49)

--gain of > 20#: OR, 1.00 (0.77,1.30)

p-trend = 0.46

Weight change between 18 and 30 years by case subjects’ age at diagnosis:

>  30 to ≤40 years

--loss of ≥10#: OR, 0.47 (0.28,0.79)

--loss of < 10# to gain of ≤10#: OR, 1.00

--gain of 10 to ≤20#: OR, 1.25 (0.91,1.71)

--gain of > 20#: OR, 1.03 (0.72,1.47)

p-trend = 0.48

>  40 years

--loss of ≥10#: OR, 0.97 (0.52,1.65)

--loss of < 10# to gain of ≤10#: OR, 1.00

--gain of 10 to ≤20#: OR, 1.16 (0.85,1.59)

--gain of > 20#: OR, 0.95 (0.64,1.43)

p-trend = 0.75

Weight change between 18 and 30 years byBRCA1/2pathogenic germline gene variant:

BRCA 1

--loss of ≥10#: OR, 0.35 (0.18,0.67)

--loss of < 10# to gain of ≤10#: OR, 1.00

--gain of 10 to ≤20#: OR, 1.29 (0.91,1.83)

--gain of > 20#: OR, 1.09 (0.73,1.62)

p-trend = 0.34

BRCA2

--loss of ≥10#: OR, 0.88 (0.35,2.23)

--loss of < 10# to gain of ≤10#: OR, 1.00

--gain of 10 to ≤20#: OR, 1.08 (0.50,2.35)

--gain of > 20#: OR, 0.77 (0.33,1.81)

p-trend = 0.70

When assessed by pathogenic gene variant and parity (0, 1, ≥2), with loss of < 10# between age 18 and 30 years as the reference group, the only significant association observed was with gain of > 10# among women withBRCA 1pathogenic germline gene variant with parity ≥2, OR, 1.44 (1.01,2.04)

Only univariate results reported in paper. Per authors, results from analyses adjusted for oral contraceptive use, smoking, oophorectomy, and parity were similar to univariate results.

558 women with BRCA 1 pathogenic germline gene variants

167 women with BRCA 2 pathogenic germline gene variants

218 women diagnosed with breast cancer within the 10-year period of questionnaire

− 170 BRCA 1

− 48 BRCA 2

Retrospective Cohort, data from HEBON study (Hereditary Breast and Ovarian Cancer Study, the Netherlands)

Weight status/ weight change

Standardized risk factor questionnaire

Questions related to body weight/weight change include weight at age 18, current weight and current height, body weight in different age periods (10-year increments starting at age 20 up to 70+)

Specifically assessed weight change in relation to menopausal status

Significant associations were not observed among the following variables in relation to premenopausal breast cancer risk among women with BRCA 1/2 pathogenic germline gene variants: body weight at age 18, BMI at age 18, current body weight, current BMI, adult weight change, and relative weight change*^,**.

Current weight (kg) and postmenopausal breast cancer risk*^,***:

< 72: 1.00

≥72: HR, 2.10 (1.23,3.59)

-No other significant associations observed for weight change variables and postmenopausal breast cancer risk.

*Data analyzed as time-varying Cox-proportional hazards model, stratified by gene and birth cohort, clustered for family, and adjusted for parity, type of menopause and history of hormone replacement therapy, and lifetime sports activity.

**Results observed were the same for weighted cohort approach analysis and unweighted analysis.

***Results are for unweighted analysis, underpowered to conduct weighted cohort approach analysis.

7516 women with BMI data

− 4401 BRCA1

− 3115 BRCA2

Total sample size for Consortium of Investigators of Modifiers of BRCA1/2 (CIMBA):

22,588 women with BRCA1 or BRCA2 pathogenic germline gene variant

14,676 BRCA1

− 7360 women with breast cancer (cases)

7912 BRCA2

− 4091 cases

Case-Control, data from CIMBA − 33 countries including 55 centers

Weight status

Questionnaire of self-reported height and weight to calculate observed BMI at date of questionnaire and during young adulthood

Included Mendelian Randomization approach: Calculated weighted genetic score for BMI and height (see paper for details)

Observed BMI and Breast Cancer Risk at Date of Questionnaire:

Per 5 kg/m² (participants/number of events):

-All participants (6964/3331): HR, 0.94 (0.90, 0.98)*

-BRCA1 (4401/2114): HR, 0.96 (0.91, 1.01)**

-BRCA2 (3115/1480): HR, 0.90 (0.84, 0.97)**

-Premenopausal (7516/2153): HR, 0.92 (0.87, 0.97)***

-Postmenopausal (3029/1389): HR, 0.97 (0.91, 1.04)***

Observed BMI and Breast Cancer Risk in Young Adulthood:

Per 5 kg/m² (participants/number of events):

-All participants (5210/2436): HR, 0.82 (0.75, 0.90)*

-BRCA1 (3134/1462): HR, 0.87 (0.78, 0.97)**

-BRCA2 (2283/1058): HR, 0.74 (0.63, 0.85)**

-Premenopausal (5417/1519): HR, 0.85 (0.78, 0.94)***

-Postmenopausal (2181/977): HR, 0.79 (0.69, 0.91)***

BMI Genetic Score and Breast Cancer Risk at Date of Questionnaire:

Per 5 kg/m² (participants/number of events):

-All participants (22,588/11,451): HR, 0.87 (0.76, 0.98)****

-BRCA1 (14,676/7360): HR, 0.88 (0.76, 1.02)**

-BRCA2 (7912/4091): HR, 0.83 (0.65, 1.05)**

-Premenopausal (22,588/7410): HR, 0.84 (0.73, 0.98)*****

-Postmenopausal (8459/3926): HR, 0.89 (0.72, 1.09)*****

*Data from fully adjusted model. Adjusted for principal components, country, birth cohort, mutation status, menopausal status, parity and age at menarche.

**Data adjusted for principal components, birth cohort, country of enrollment, menopausal status.

*** Data adjusted for principal components, birth cohort, country of enrollment, mutation status.

****Data adjusted for principal components, birth cohort, country of enrollment, menopausal status, mutation status.

*****Data adjusted for principal components, birth cohort, country of enrollment, menopausal status.

Data for other multivariable adjustments and height are available in the paper.

104 women with BRCA1/2 pathogenic germline gene variants

67 BRCA1

37 BRCA2

Retrospective cohort of Ashkenazi Jewish women, data from 12 participating cancer centers in the greater New York City area

Weight status

Physical activity

Data collection method not provided in detail

Weight status was inquired at menarche and age 21

Physical activity behavior was inquired during adolescence

Normal weight status (per BMI) at menarche (p = 0.017) and age 21 (p = 0.021) was associated with breast cancer onset at an older age*.

Engagement in physical activity as a teenager was associated with breast cancer onset at an older age (p = 0.034)*.

*Data adjusted for decade of birth of the proband.

443 matched pairs of women with BRCA1/2 pathogenic germline gene variants

Case-control, data from 80 participating centers in 17 countries

Physical activity

Nurses’ Health Study II Physical Activity Questionnaire

Standardized questionnaire including questions related to family history, medical and personal history, reproductive, hormonal and lifestyle factors

Total physical activity (moderate + vigorous) and vigorous physical activity alone was not significantly associated with breast cancer risk in adolescence (ages 12–17), young adulthood (ages 18–34), and overall (ages 12–34). Significant associations were not observed when assessed by menopausal status (i.e. pre- or postmenopausal) at breast cancer diagnosis*.

Significant association was not observed for moderate physical activity among all age groups when assessed for the total sample, by postmenopausal status at breast cancer diagnosis, and among the young adulthood and overall (young adulthood + adolescence) for premenopausal status at breast cancer diagnosis. The only significant association observed was for adolescent physical activity and premenopausal at breast cancer diagnosis (see below for data)*.

Moderate physical activity in adolescence and premenopausal breast cancer risk*:

≤6.75 MET-hrs/week: 1.00

> 6.75 and ≤ 15.75 MET-hrs/week: HR 1.04 (0.70,1.53)

> 15.75 and ≤ 25.88 MET-hrs/week: HR 1.48 (0.94,2.32)

> 25.88 MET-hrs/week: HR 0.62 (0.40,0.96)

p-trend = 0.01

*Data adjusted for number of children, current BMI, oral contraception use, tobacco consumption, and history of oophorectomy.

558 women with BRCA1 pathogenic germline gene variants

167 women with BRCA2 pathogenic germline gene variants

218 carriers diagnosed with breast cancer within the 10-year period of questionnaire

− 170 BRCA1

− 48 BRCA2

Retrospective Cohort, HEBON study (Hereditary Breast and Ovarian Cancer Study, the Netherlands)

Physical activity

Standardized risk factor questionnaire

Questions related to physical activity behavior include: type of sport, number of hours spent per week, ages at which it was practiced. Questions were specific to activities performed for at least 6 months for at least 1 h/week.

Significant associations were not observed between the following activity variables and breast cancer risk when never engaging in lifetime sports activity was the reference group: Mean MET hours/week (low, < 11.0; medium, 11.0–22.7; high, ≥22.7), Mean hours/week (low, < 2.0; medium, 2.0–3.3; high, ≥3.3), Number of active years (<  9 years, 9–19 years, ≥19 years).

Lifetime sports activity and breast cancer risk *:

Mean MET hours/week

--low (< 11.0): 1.00

--medium (11.0–22.7): HR, 0.59 (0.36,0.95)

--high (≥22.7): HR, 0.77 (0.48,1.24)

p-trend = 0.494

Significant associations were not observed for Mean hours/week and number of active years when the lowest category was used as the reference category.

Lifetime sports activity before age 30 and breast cancer risk*:

Mean MET hours/week

--low (< 11.0): 1.00

--medium (11.0–22.7): HR, 0.60 (0.38,0.96)

--high (≥22.7): HR, 0.58 (0.35,0.94)

p-trend = 0.053

Significant associations were not observed for Mean hours/week and number of active years when the lowest category was used as the reference category.

Significant associations were not observed for activity variables when never engaging in lifetime sports activity was the reference group.

Lifetime sports activity after age 30 and breast cancer risk*:

Mean MET hours/week

--never engaging in activity: 1.00

--low (< 11.0): HR, 0.55 (0.34,0.90)

--medium (11.0–22.7): HR, 0.70 (0.44,1.14)

--high (≥22.7): HR, 0.68 (0.43,1.09)

p-trend = 0.157

Mean hours/week

--never engaging in activity: 1.00

--low (< 2.0): HR, 0.53 (0.32,0.86)

--medium (2.0–3.0): HR, 0.80 (0.47,1.36)

--high (≥3.0): HR, 0.66 (0.42,1.04)

p-trend = 0.135

Number of active years

--never engaging in activity: 1.00

-- < 5: HR, 0.52 (0.32,0.85)

--5-11: HR, 0.78 (0.48,1.26)

-- ≥ 11: HR, 0.64 (0.39,1.03)

p-trend = 0.119

Sports activity

--never: 1.00

--ever: HR, 0.63 (0.44,0.91)

Significant associations were not observed for Mean hours/week and number of active years when the lowest category was used as the reference category.

Recent sports activity and breast cancer risk by time windows:

1 year

Mean hours/week

--low (< 2.0): HR, 0.48 (0.26,0.87)

--medium (2.0–3.0): HR, 0.90 (0.55,1.47)

--high (≥3.0): HR, 0.90 (0.58,1.40)

Significant associations were not observed for Mean MET hours/week or percent active years.

2 years

Mean hours/week

--low (< 2.0): HR, 0.49 (0.29,0.85)

--medium (2.0–3.0): HR, 0.89 (0.52,1.50)

--high (≥3.0): HR, 0.94 (0.61,1.44)

Significant associations were not observed for Mean MET hours/week or percent active years.

5 years

Mean MET hours/week

--low (< 11.0): HR, 0.64 (0.42,0.98)

--medium (11.0–22.7): HR, 0.91 (0.56,1.50)

--high (≥22.7): HR, 0.92 (0.57,1.50)

Significant associations were not observed for Mean hours/week or percent active years.

10 years

Significant associations were not observed for this time window.

*Data adjusted for use of oral contraceptives, parity, menopausal status, hormone replacement therapy use, age-specific BMI, BMI at age 18, alcohol consumption, occupational activity. Mean METhours/week and mean hours/week also adjusted for number of active years. Number of active years also adjusted for mean METhours/week.